Rapid increase in lymphogranuloma venereum in men who have sex with men , United Kingdom , 2003 to September 2015

T Childs 1 , I Simms 1 , S Alexander 2 , K Eastick 3 , G Hughes 1 , N Field 1 4 1. HIV and STI Department, Public Health England Health Protection Services, Colindale, United Kingdom 2. Sexually Transmitted Bacteria Reference Unit, Public Health England Reference Microbiology Services, Colindale, United Kingdom 3. Scottish Bacterial Sexually Transmitted Infections Reference Laboratory, Edinburgh Royal Infirmary, Edinburgh, United Kingdom 4. Department of Infection and Population Health, University College London, London, United Kingdom


Testing protocol and data sources
Rectal, genital or urine samples from patients with symptoms compatible with LGV and diagnosed with Chlamydia trachomatis in England, Wales and Northern Ireland, and their sexual contacts with C. trachomatis, are submitted by local laboratories for LGV typing to the Public Health England national reference laboratory, the Sexually Transmitted Bacteria Reference Unit (STBRU) in London.Scottish samples from symptomatic patients are similarly submitted to the Scottish Bacterial Sexually Transmitted Infections Reference Laboratory (SBSTIRL) in Edinburgh.In addition, since 2007, samples from asymptomatic HIV-positive MSM diagnosed with C. trachomatis are also sent to SBSTIRL.During the reporting period, (January 2003 to September 2015) STBRU and SBSTIRL diagnoses accounted for all confirmed LGV cases diagnosed in the UK.Duplicates were excluded and repeat diagnoses within 42 days (used to define a single episode of care across surveillance systems in England) were counted as a single episode [1].
Additional data, including sexual behaviour and clinical information, were collected through nationally coordinated enhanced surveillance of LGV in the UK between 2004 and 2010 [2].However, the most recent laboratory dataset lacks this enhanced information.We therefore matched laboratory reports for LGV diagnoses to the Genitourinary Medicine Clinic Activity Dataset (GUMCADv2), an anonymised patient-level electronic dataset, for all cases in England in 2014.All commissioned sexual health services in England have been mandated to return GUMCADv2 data, which include patient characteristics, other sexually transmitted infection (STI) diagnoses and services provided, since 2009 [3].A code for LGV diagnosis was introduced in 2011, such that most cases are now reported through GUMCADv2.
LGV diagnoses in female or heterosexual male patients in GUMCADv2 were excluded from this analysis because they may represent miscoding (six cases).

Background
LGV is a STI caused by the invasive L serovars of C. trachomatis.An outbreak among MSM was first reported in the Netherlands in 2003 and LGV outbreaks in MSM have since been reported by many high-income countries, with HIV co-infection being a common feature [4][5][6][7][8].Most reported infections are rectal, and the most common presentation is proctitis, associated with rectal pain, discharge and bleeding [5,9].Some cases of LGV, particularly in the beginning of the outbreak, were misdiagnosed as Crohn's disease [10,11].The complications of incorrectly treated or untreated LGV are serious, particularly for immunocompromised individuals, including genital ulcers, fistulas, rectal strictures and genital elephantiasis.
LGV acquisition is associated with concurrent STI infections and reported behaviours such as condomless anal intercourse, fisting and the use of sex toys [2].

Discussion
Since 2003, LGV diagnoses have increased over a twelve-year period, with a steep rise since 2012, to reach unprecedented levels in 2015, suggesting high levels of ongoing transmission.Infection remains concentrated in white MSM living in London, many with HIV co-infection and other STIs.LGV: lymphogranuloma venereum.The strength of this study lies in the unique national dataset of laboratory confirmed LGV diagnoses, which we linked to clinical data for the most recent complete year.During the study period, LGV testing was only recommended for symptomatic patients and LGV contacts.However, a large multicentre case-finding study in the UK found that 27% of LGV infections were detected in patients without rectal symptoms, suggesting that infections might be missed [12,13].Patients might also be treated presumptively without LGV testing or never seek care, leading to further underestimation of LGV cases in the population.Other external factors might have influenced requests for LGV testing and the trends observed.These include the introduction of charging laboratories for LGV diagnostic testing (April 2014) and increased awareness of LGV among clinicians and microbiologists, which might have led to increased testing.We lack data on symptoms associated with LGV infection in this surveillance dataset.
The increase in LGV diagnoses coincides with upward trends in other STIs and sexually transmissible enteric infections among MSM, such as gonorrhoea, infectious syphilis, verotoxin-producing Escherichia coli and Shigella [14][15][16][17].Contributory factors might include sexual networks with high rates of partner change facilitated by social media networking sites and chemsex, and we know that HIV-positive MSM who engage in HIV sero-adaptive behaviours and have condomless sex with other HIV-positive men remain at particularly high risk [15][16][17].Our data suggest a need for a strengthened public health response to raise awareness of LGV among clinicians and patients.Social media might offer a means of rapid, low-cost dissemination of public health messages, with an additional benefit of facilitating user interaction [18].
There are also important implications for clinical management and microbiologists, including the need for robust partner notification, testing and treatment, and testing of HIV-positive MSM with asymptomatic C. trachomatis infection for LGV [19].
LGV treatment requires an extended course of antibiotics over non-LGV C. trachomatis, such that some patients may unnecessarily receive multiple courses of antibiotics; this risks introducing selection pressure for antimicrobial resistance [20].Clinicians should maintain a high index of suspicion for LGV in HIV-positive MSM with rectal symptoms, and consider treating presumptively for LGV according to national guidelines [21,22].

Figure
Figure Number of cases diagnosed with lymphogranuloma venereum, per quarter, United Kingdom, 2003 to end September 2015 (n=4,124)

Table
Characteristics of MSM patients from STBRU laboratory reports matched to GUMCADv2 dataset, United Kingdom, 2014 (n=434) Of 677 positive LGV test results in 2014, 440 were successfully matched to GUMCADv2.Six of these were excluded because they were recorded as female or heterosexual in GUMCADv2, leaving 434 matched patients.All other STI diagnoses were new except for 14 recurrent cases of genital herpes and 17 recurrent cases of genital warts.
a b For patients with information available.c